I think this is my favorite passage from Monk's paper:
One reason (not mentioned by Hacker) for being suspicious of the tendency to regard Wittgenstein's later philosophy as a kind of therapy is that, although in this context, the word "therapy" is used as the opposite of "theory," in almost all other contexts it is assumed that a form of therapy is founded upon and shaped by a particular theory. To psychoanalytic therapy, there corresponds Freudian theory, to Gestalt therapy there corresponds Gestalt theory, to "primal therapy" there corresponds Janov's theory of repressed pain, and so on. To call Wittgenstein's later work "therapy" is not, therefore, necessarily to assume that it does not express a theory; on the contrary, it might well invite the question of what theory this therapy is based on.I suspect I have not been as attentive to this point as I should have been. It misses Wittgenstein's point to approach all philosophy as "sick" and in need of treatment, for one can rightly take up such a position only if one comes in already knowing a great deal about philosophy -- so that one can recognize illness for illness and not health. But one can't have this sort of knowledge; philosophy doesn't have the sort of unity a species of animal does, so there cannot be the same sort of distinction between illness and health you can draw in medicine. And anything a philosopher can do, they can do; it is not the Wittgensteinian's place to forbid them anything. It is important not to take the position of a doctor treating a patient when trying to "lead words from their metaphysical back to their everyday use"; doctors have knowledge that justifies their taking the stance they take. There is nothing analogous that could allow a Wittgensteinian to identify a "metaphysical use" of a word. All that there is to be done is to either learn what sense they attach to their words, and then proceed from there (beyond philosophy), or to come to a shared recognition that their words have no clear sense to them.
I think this is the irony of TLP 6.53
The right method of philosophy would be this: To say nothing except what can be said, i.e. the propositions of natural science, i.e. something that has nothing to do with philosophy: and then always, when someone else wished to say something metaphysical, to demonstrate to him that he had given no meaning to certain signs in his propositions.when considered beside TLP 4.11
The totality of true propositions is the total natural science (or the totality of the natural sciences).
Reading 6.53 by itself, it might appear that Wittgenstein is advocating a sort of naturalism or positivism: we should take the natural sciences as a model for all thinking, and avoid philosophizing. This is a sort of perspective which it isn't hard to find proponents of; Dawkins comes to mind. But when a Dawkins praises science, he has in mind things like physics, chemistry, and biology. You can identify something as a science (in this sense) by its content. It is opposed to things like mathematics, history, and philosophy. (To say that life should be approached in a "mathematical spirit" or a "historical spirit" is very different from saying, with the naturalist, that life should be handled scientifically.) But this isn't how the Tractatus delimits "natural science": 4.11 tells us that by "the totality of natural science" Wittgenstein means merely "the totality of true propositions". You can't tell from the content of a proposition whether or not it's a scientific one; the mere fact of it having content settles all that one is concerned with, if one is trying to follow the advice of 6.53. To say nothing but the propositions of natural science is just to not talk nonsense.
But "Don't talk nonsense" isn't something one can accomplish by deciding to do it; it is a key thought of Wittgenstein's, both early and late, that it is easy to start speaking nonsense without realizing one has begun doing so. Even if one is aware that this is easy to do, one still slips into it from time to time. One can't follow "the right method of philosophy" because of this.
An analogous problem affects the other half of 6.53. It might seem that we can recognize someone's wish "to say something metaphysical" by the content of what they're trying to say. A "metaphysical" statement would be something about God, the soul, the nature of Reality, Being -- the sorts of things one finds badly handled in the "Metaphysics" section of a bookstore, and handled not much better in philosophy departments. But this isn't how the Tractatus thinks of metaphysics, either. Wittgenstein doesn't give us an account of what metaphysics is in the Tractatus (which is worth noting by itself), but the method he prescribes for dealing with it in 6.53 is clearly an echo of 5.4733:
Every possible proposition is legitimately constructed, and if it has no sense this can only be because we have given no meaning to some of its constituent parts. (Even if we believe that we have done so.)which is a comment on 5.473, where we find
A possible sign must also be able to signify. Everything which is possible in logic is also permitted. (“Socrates is identical” means nothing because there is no property which is called “identical”. The proposition is senseless because we have not made some arbitrary determination, not because the symbol is in itself unpermissible.)-- and looking at these passages together, it seems clear that "metaphysics" is just a label for nonsense.
So to follow the advice of 6.53, we would need to be able to tell when someone wanted to say something nonsensical. But we cannot tell when this is happening, as a rule: nonsense can slip us by. Nor are there any particular topics which we can know a priori will constitute a slide into "metaphysics" if they come up. TLP 5.557:
The application of logic decides what elementary propositions there are. What lies in its application logic cannot anticipate.-- and since, in the Tractatus, all propositions are truth-functions of elementary propositions, logic cannot anticipate what propositions there are. And as a proposition is just what has sense, we cannot anticipate what will have sense. How, then, are we to tell whether something has a sense or not?
To answer this, I think it helps to look at TLP 3.326:
In order to recognize the symbol in the sign we must consider the significant use.and 3.227
The sign determines a logical form only together with its logical syntactic application.and 3.363
What does not get expressed in the sign is shown by its application. What the signs conceal, their application declares.and, as a terminological reminder, 3.2:
The sign is the part of the symbol perceptible by the senses.Our trouble, in trying to follow the advice of 6.53, is to determine from the signs (sounds, ink-marks) someone has given us whether or not we are dealing with "metaphysics". But nothing in a sign tells us what symbol we are to recognize in it: 3.21
Two different symbols can therefore have the sign (the written sign or the sound sign) in common—they then signify in different ways.and 3.22 (my emphasis)
It can never indicate the common characteristic of two objects that we symbolize them with the same signs but by different methods of symbolizing. For the sign is arbitrary. We could therefore equally well choose two different signs and where then would be what was common in the symbolization?So what we need to know, when presented with some signs, cannot be "read off" from the signs themselves. The sign does not tell us whether it is being used metaphysically, and neither does logic.
Consider now 4.002:
Colloquial language is a part of the human organism and is not less complicated than it. From it it is humanly impossible to gather immediately the logic of language.... The silent adjustments to understand colloquial language are enormously complicated.which is followed by 4.003
Most propositions and questions, that have been written about philosophical matters, are not false, but senseless. We cannot, therefore, answer questions of this kind at all, but only state their nonsensicality. Most questions and propositions of the philosophers result from the fact that we do not understand the logic of our language.It seems that here we get another version of the advice of 6.53: we are to state the nonsensicality of propositions which philosophers have often tried to put forth (which are, one presumes, "metaphysics"). But there is no temptation to think that this can simply be taken up as a "method" here: we are told that understanding colloquial language is "enormously complicated" and that it is humanly impossible to immediately gather the logic of such language.
To recognize nonsense as nonsense, we need to consider the significant use (sinnvollen Gebrauch) of a sign and recognize the symbol in it (recognize its application). If this is impossible, then we're dealing with nonsense -- but a verdict that this is impossible can only be a pessimistic induction over failed attempts to "consider its signifcant use". Logic always leaves open that we have simply, thus far, missed how the sign is being used.
So if we try to take 6.53's advice and follow "the only strictly correct method" in philosophy, we will quickly find that "there easily arise the most fundamental confusions (of which the whole of philosophy is full)": we treat as nonsense what is not, and as not nonsense what is; we take ourselves to be in possession of a method where there cannot be one. For we do not understand ourselves or each other as well as 6.53 makes it seem like we can: 5.5563
(Our problems are not abstract but perhaps the most concrete that there are.)